Investment history is a recurring cycle of euphoria and collapse, and China’s current obsession with 'hard tech' and AI appears to be the latest chapter. From the Nasdaq’s 2000 dot-com crash to the 2015 'Internet+' mania in China’s A-share market, the signs of a peak are historically consistent. Today, the STAR Market—encompassing AI and 'new quality productive forces'—exhibits all the classic symptoms of a bubble, though it is dressed in a more sophisticated nationalistic narrative.
Valuation metrics for the STAR Market have reached startling extremes. Trading at a price-to-earnings (PE) ratio between 80 and 100 times projected 2026 profits—and as high as 200 times 2025 levels—these figures significantly eclipse the Nasdaq’s 40x multiple. This decoupling from fundamental reality mirrors the height of the 2015 speculative frenzy, where 'future imagination' replaced current earnings as the primary valuation anchor. While the mantra in 2015 was that the internet would change the world, today’s conviction is that AI will change everything.
The current bubble is also more deeply entrenched due to a shift in participating players. Unlike the retail-driven madness of 2015, today’s surge is propelled by an 'institutional consensus' that fuses grand national narratives with industrial capital. When a systemic belief takes hold—in this case, that AI is essential for national strength—risk is systematically underestimated. This has led to the 'rationalization of losses,' where investors forgive missing earnings and view deep deficits as necessary investments in a technological moat.
Structural vulnerabilities are particularly visible in the hardware supply chain. Leading A-share optical module firms, such as Zhongji Innolight, have seen their market caps swell to 1.5 trillion yuan despite acting largely as mid-stream processors. These companies remain dependent on foreign core components, such as chips from Lumentum, making them easily replaceable in the global architecture. As industry-wide capacity expands, profit margins are poised to contract, yet investors continue to project current growth rates indefinitely into the future.
Historical precedents, such as Warren Buffett’s 1999 warnings at Sun Valley, suggest that while innovation changes the world, it rarely makes early speculators rich. In 2000, 90% of the companies at the peak of the Nasdaq eventually saw their values drop to zero, not because the internet failed, but because their technical paths or profit timelines were flawed. We are approaching a similar inflection point where any commercial disappointment or price war in AI models could puncture the prevailing optimism.
For investors seeking to survive the eventual correction, the priority must shift from 'silicon-based' narrative plays to 'carbon-based' essentials like consumption and healthcare. Holding cash and focusing on core assets with robust cash flows—such as Tencent or Alibaba—remains the only proven strategy for traversing a boom-and-bust cycle. As the gap between market narrative and operational reality reaches its limit, liquidity and diversification will distinguish those who survive from those who are wiped out by the burst.